Intro

2020 is a special year, and this year in US, policeman shot African American for almost no reason triggered people’s anger again and again. This always reminds us that racial justice is a very common and serious topic in the US. In this report, we will understand racial justice more comprehensively. We will not only just have a basic concept about racial justice, but also we will know how racial justice expresses in different fields. We will achieve this by doing analysis on data sets about COMPAS Recidivism Racial Bias, Unemployment Rate by Race and Ethnicity, Juvenile Arrests and US Police Shootings.

Summary Info

From the data set of inmates, we can see African-American was the ethnicity with most inmates. Among these inmates, 3030 were considered with a high risk level, and this is 5 times of the numbers of white inmates who were considered with a high risk level. This wide gap between the statistics of two ethnic groups shows that black inmates might be wrongly considered as high risk because of social injustice. From the data set of juvenile arrests, we can see BLACK has the most juveniles arrested, and the main reason for that is just ISSUANCE OF WARRANT, which seems to be unreasonable. It shows the stereotype people have to African Americans. The data set of unemployment rate also shows that African American is the group with the highest unemployment rate. From all of these data sets, we can see the seriousness of racial justice in the US.

Summary table

Here is a summary table of the total number or frequency of charges against each race and shows the percentage of charges that were treated as a felony for each year from 2013 - 2017

Date Race Frequency Convicts Felony Percentage
2013 BLACK 16840 4069 24.16
2013 WHITE HISPANIC 3558 480 13.49
2013 WHITE 667 91 13.64
2013 BLACK HISPANIC 134 29 21.64
2013 ASIAN PACIFIC ISLANDER 52 7 13.46
2013 UNKNOWN 13 4 30.77
2013 AMERICAN INDIAN / ALASKAN NATIVE 1 0 0.00
2014 BLACK 13932 3784 27.16
2014 WHITE HISPANIC 2937 508 17.30
2014 WHITE 525 101 19.24
2014 BLACK HISPANIC 117 25 21.37
2014 ASIAN PACIFIC ISLANDER 48 9 18.75
2014 UNKNOWN 6 0 0.00
2014 AMERICAN INDIAN / ALASKAN NATIVE 4 2 50.00
2015 BLACK 11143 3231 29.00
2015 WHITE HISPANIC 2699 576 21.34
2015 WHITE 495 105 21.21
2015 BLACK HISPANIC 107 29 27.10
2015 ASIAN PACIFIC ISLANDER 56 9 16.07
2015 UNKNOWN 12 2 16.67
2015 AMERICAN INDIAN / ALASKAN NATIVE 4 1 25.00
2016 BLACK 8514 2813 33.04
2016 WHITE HISPANIC 1795 482 26.85
2016 WHITE 349 68 19.48
2016 BLACK HISPANIC 78 25 32.05
2016 ASIAN PACIFIC ISLANDER 59 19 32.20
2016 UNKNOWN 11 1 9.09
2016 AMERICAN INDIAN / ALASKAN NATIVE 3 0 0.00
2017 BLACK 2096 649 30.96
2017 WHITE HISPANIC 439 147 33.49
2017 WHITE 87 21 24.14
2017 ASIAN PACIFIC ISLANDER 23 5 21.74
2017 BLACK HISPANIC 18 5 27.78
2017 UNKNOWN 4 0 0.00

Each year’s frequency was listed in a decreasing way. From this approach, we can see from 2013 to 2017, African American always had the most juveniles arrested. The number of arrested juveniles of black people was always at least 3 times of the number of the arrested juveniles of the race with the second highest frequency. African American was also the ethnic group with the highest felony percentage.

Three charts

Chart 1:

Chart 1 shows the number of individuals involved in police shootings from 2015-2020 based on their race. The chart also shows whether the individual was aggressive, other or undetermined.

  1. This chart shows that the majority of fatal police shootings involved the black, hispanic, and white people. Unsurprisingly, the majority of the victims had violent intentions and blacks had the highest percentage of aggressive intent at 67.11% followed closely by whites at 66.14%.
  2. The difference in victims between the other 3 races (including other as one) is quite slim however the length visualisation helps realise that Asians are affected more than the Native Americans.

Chart 2:

Chart 2 is about compas dataset shows the decile scores by the number of people falls in that decile colored by different races. Decile score: 1-4 Low Recidivism Level, 5-10 Medium and High Level(or High supervision level)

  1. This chart shows that scores for Caucasian defendants were skewed toward lower-risk categories, while African-American defendants were evenly distributed across scores. Notice that all the defendants in this chart are those who had not been arrested for a new offense or who had recidivated within two years.
  2. This chart also shows for the defendants who had not recidivated within two years, the number of African-American defendants who were categorized as the highest-risk(decile score is 10) is about the same as the number of Caucasian defendants who were categorized as the medium-risk(about score 5), or the same as the number of Hispanic or other races defendants who were categorized as the lowest-risk(about score 2.5). Notice that the number is the same but how the campas scores of recidivism are different.

Chart 3:

Chart 3 is about the Compas dataset and shows if the race will be a factor that influences the COMPAS score. When the number on the y axis is higher than 0, it means being this particular race will have more chance of being categorized as high-supervision (scored as medium or high, or >= 5), and vice versa.

  1. This chart shows that when other variables like age or sex are controlled as the same, the race will be a factor for COMPAS scoring, and the difference for each races is significant.
  2. From this chart, being an African-American will have the highest chance of being scored as high-risk while being Hispanic(apart from the category “Other”) will have the lowest.